Return Series (money-weighted)

The Return series is EquityTruth's money-weighted equity curve — the one that answers "what actually happened to the money in this account, as a percentage." (Money-weighted means each trade's percentage is measured against your balance at that moment, so the more money in the account, the smaller each trade's % impact.) Most equity metrics (CAGR, Sharpe, drawdown) are computed on it by default. Its defining trait: deposits and withdrawals move it, because they change the balance your return is measured against. That isn't a flaw — it's the right answer to "what did my money actually do." It only becomes the wrong tool when you're judging the strategy in isolation (independent of when you funded it); for that, use the TWR series.

How it's built

base_i  = cumulative_cashflows_before_i + cumulative_net_pnl_before_i
r_i     = net_pnl_i / base_i
Return  = (∏(1 + r_i) − 1) × 100%

Because the return of each trade is divided by the balance at that moment, a deposit that doubles the balance halves the percentage impact of every subsequent trade — that's the money-weighting in action.

Close-to-close vs low-to-low

Every equity series is sampled once per day in one of two ways — and which one a metric uses is decided by the engine, not by you. The two names are parallel: each is [point]-to-[point], differing only in which point of the day they read.

• close-to-close — one value per day, taken at the day's close, the next day's return measured from one close to the next. The everyday end-of-day path.
• low-to-low — one value per day, taken at the day's intraday low (the worst point reached within the day), walked from one day's low to the next day's low. The worst-point path.

The rule: metrics that include a measure of risk use low-to-low; metrics not directly about risk use close-to-close. Risk lives in the intraday dips, so anything that judges downside, swings, or the shape of the return distribution reads the daily low rather than the polite close:

• low-to-low (risk-aware) — Sharpe, Sortino, Volatility, VaR, CVaR, Omega, and the distribution-shape metrics (skewness, kurtosis, tail ratio). They care about how far returns swing and how heavy the loss tail is.
• close-to-close (not risk) — total return, CAGR, win-days %, time under water, stability. They count days, measure the final level, or track the trend — the close is enough.

Both are still one observation per day (low-to-low just takes the day's worst point instead of its close), which is why annualization uses ≈252 observations/year — the number of trading days. The honest caveat: the intraday low is reconstructed from minute data, so its accuracy depends on the quality of that reconstruction.

Max drawdown is the one hybrid: its high-water mark is taken from the close (the intraday high never sets a new peak), but the depth below that peak is measured at the intraday low — the most conservative read on both ends.

Related series

The Return series is one of three equity curves. The TWR series removes cashflows; the P&L series drops percentages for absolute dollars. The per-trade Return axis is a different thing — it's a single trade's % result, not a cumulative curve.